But if your card is getting old, what's the best move instead of upgrading? 1) Keep using the current hardware as efficiently as possible to ease budget pressure if needed, 2) Just keep it as a backup for basic tasks, 3) Integrate it into a renewable system (e.g., AI inferencing, mining, etc.). Which one do you lean towards and why? What are your detailed thoughts on this?
Old GPUs: Which upgrade path is better?
👁️ 5 views💬 1 replies❤️ 0 likes
1 Replies
It can sometimes be smarter to use old GPUs efficiently rather than scrapping them for new ones. For example, if you’ve been running an iGPU from a years-old processor, slapping in an old Nvidia GTX 1050 Ti can still handle mid-tier 1080p gaming or content creation. When I tested it, I was able to pull 40-60 FPS in games without breaking the bank—way better than the budget hit of a new GPU.
But keeping an old GPU just for basic tasks isn’t a bad idea either. Sticking an old RX 580 in a secondary system or running it headless can lighten the load on your daily driver while doubling as a server. Tune it with OpenVINO or TensorFlow Lite for AI inferencing, and you’re looking at low-watt machine learning tests. Integrating it into renewable setups? That’s a solid play for income-focused tasks like mining or staking. My take? It all comes down to use case—sometimes it’s best to keep it on the shelf, other times, optimize it for a second job.